The lighting theory and structure of light-emitting diode (LED) is different from that of conventional lighting source. LED has advantages as a low power loss, a long life-time, no need for warming time, and fast responsive time. Moreover, it is small, shockproof, suitable for mass production, and highly compatible with application demand for a tiny or array-type element, so LEDs are widely adopted in the market. For example, LEDs can be used in optical display apparatus, laser diodes, traffic lights, data storage devices, communication devices, illumination devices, medical devices, and so on.
As shown in FIG. 1, a conventional light-emitting array includes: an insulative sapphire substrate 10; a plurality of light-emitting stacks 12 formed on the insulative sapphire substrate 10 and including a p-type semiconductor layer 121, an active layer 122, and an n-type semiconductor layer 123. Because the sapphire substrate 10 is insulative, the light-emitting stacks can be insulated from each other by forming trenches therebetween with etching processes. Furthermore, after partially etching the plurality of light-emitting stacks 12 to the n-type semiconductor layer 123, a first electrode 18 is formed on the exposed area of the n-type semiconductor layer 123, and a second electrode 16 is formed on the p-type semiconductor layer 121. Metal wires 19 are then provided to selectively connect the first electrode 18 and the second electrode 16 to connect the plurality of light-emitting stacks 12 in parallel or series configuration.